Workplace Sexual Harassment in Two General Hospitals in Taiwan: The Incidence, Perception, and Gender Differences

Transcription

1 56 J Occup Health, Vol. 54, 2012 J Occup Health 2012; 54: Journal of Occupational Health Field Study Workplace Sexual Harassment in Two General Hospitals in Taiwan: The Incidence, Perception, and Gender Differences Liang-Jen Wang 1 3, Chih-Ken Chen 1, 3, Yi-Chen Sheng 4, Pei-Wen Lu 5, Yi-Ting Chen 1, Huei-Jun Chen 6 and Jyh-Sheng Lin 7 1 Department of Psychiatry, Chang Gung Memorial Hospital, 2 College of Public Health, National Taiwan University, 3 Chang Gung University School of Medicine, 4 Department of Psychiatry, Tainan Hospital, Department of Health, 5 Department of Social Work, Chang Gung Memorial Hospital, 6 Chang Gung University of Science and Technology and 7 Institute of Education and Center of Teacher Education, National Taiwan Ocean University, Taiwan Abstract: Workplace Sexual Harassment in Two General Hospitals in Taiwan: The Incidence, Perception, and Gender Differences: Liang-Jen WANG, et al. Department of Psychiatry, Chang Gung Memorial Hospital at Keelung, Taiwan Objectives: The aims of this study were to examine sexual harassment (SH) among hospital staffs in Taiwan, in terms of three-month incidence rate, the frequency of each type and the perception of SH, perpetrated by coworkers, patients and patients families and to investigate the gender differences for these issues. Methods: The subjects were employees at two general hospitals in Taiwan. The self-administered Hospital Sexual Harassment Questionnaire was sent to eligible staff, and the voluntary respondents answered the questionnaire anonymously. There were 536 respondents available for analysis, with an overall response rate of 43.4%. Results: The three-month incidence rates of SH by coworkers, patients, and patients families in our study population were 2.4, 4.3, and 1.7%, respectively. Telling sexual jokes was the most common type of SH. The males had greater opportunities to be exposed to porn videos by their coworkers. The females were more frequently exposed to sex jokes and remarks made by patients and their family members and unwanted physical touching by patients in the workplace. There were significant differences with regard to the perception of sex jokes and sexually explicit verbal descriptions as SH or not between genders. Conclusions: The information in this study can be a helpful reference for administrators in hospitals when they are establishing education plans and policies. It might be possible to prevent sexual harassment and misunderstandings Received Mar 15, 2011; Accepted Oct 26, 2011 Published online in J-STAGE Dec 10, 2011 Correspondence to: J-S. Lin, Institute of Education and Center of Teacher Education, National Taiwan Ocean University 2 Pei Ning Road, Keelung, Taiwan, 20224, R.O.C. ( linjyhsheng@gmail. com) between genders and to further avoid the negative impact on the emotional well-being of workers in hospitals. (J Occup Health 2012; 54: 56 63) Key words: Coworker, Family, Gender, Hospital, Patient, Sexual harassment Sexual harassment (SH) is generally defined as any unwanted or unreciprocated verbal or physical sexual advance that is offensive to the person involved, and causes that person to feel threatened, humiliated or embarrassed 1, 2). SH in the workplace has become an issue of increasing concern globally in the past decades 3, 4). Hospital workers, who have to perform their daily routines with their coworkers, patients and patients families, have been reported to experience more harassment than other categories of work 5). Workers in hospitals who experienced SH easily suffer from emotional distress and unsafe feelings toward the workplace, and this might result in negative effects on the quality of patient care 6, 7). From various international studies, estimates of the annual prevalence rates of hospital staff who have experienced SH vary from 0.7 to 9.5% 8, 9). Empirically, nurses are the most vulnerable population to SH among hospital workers 10, 11). Numerous studies have reported a high prevalence of SH among nurses during their careers, ranging from 30 to 97% 12). For female doctors, SH has also been commonly reported to be as high as 77% among family physicians 13), 58.3% among residents 14) and 59% among medical students 15). In a previous study in Taiwan, the one-year prevalence of SH was 9.5% among the workers in a psychiatric hospital 9). Except for nurses and doctors, there have been relatively few studies investigating SH among the other specialties of hospital

2 Liang-Jen WANG, et al.: Sexual Harassment in General Hospitals 57 staff 16, 17). As to the issue of the perpetrators of SH, physicians and other coworkers have been reported as the most common source of SH 4, 18). Some literature has also demonstrated that the most common perpetrators of all types of SH were patients, followed by doctors, other work colleagues and visitors or patients relatives 12, 19). A broad type of behaviors can be viewed as SH. Traditionally, types of SH were categorized into quid pro quo and hostile environment 16). Fitzgerald et al. (1995) established a three-dimensional model of SH composed of gender harassment, unwanted sexual attention, and sexual coercion 20). Based on a hierarchical concept, SH has been categorized in sequence from mild (e.g., sex jokes or teasing remarks) to moderate (e.g., physical touch or repeated invitation) to severe (e.g., attempt to have sex) 21). In general, the minor types of SH were more commonly seen than the severe ones 6, 13, 19), and sex jokes seemed to be the most prevalent type of SH in hospitals. Nevertheless, the recognition and perception of SH were also diverse between genders and different cultures 19, 22, 23). For example, gender-related jokes could be perceived as either sexual humor or sexual harassment 24). It is an important issue when determining the subjective meanings of sexually related behavior to avoid unnecessary misunderstandings or potential lawsuits. As the majority of nurses are female in most countries, much of the related literature studies have investigated SH focusing on the female workers in hospitals. In recent years, there has been an increase in interest by investigators to analyze gender differences regarding SH among nurses 19, 23). The related literature has shown that SH is frequently experienced by both men and women in hospitals 15, 25). Bronner (2003) reported that female nurses and nursing students were generally more exposed to mild and moderate types of SH; however, male nurses and nursing students were more exposed to severe types of SH. doctors experienced unwanted physical advances most commonly, and males experienced unwanted attention most commonly 14). Fiedler (2000) reported that there were significant perception differences between men and women in nursing towards each type of SH. Nevertheless, studies that have examined gender differences among hospital workers, except for nurses, are still scarce. As a result, whether the results could be generalized to other populations in hospitals is still unclear. Sexual harassment, as experienced by many hospital staff, can influence the quality of medical services they offer; as such, it can eventually affect population health. Therefore, it is essential and urgent to evaluate SH among hospital workers. However, the current literature in Taiwan has not revealed the issues concerning the types and perpetrators of SH 9, 26, 27), and studies that have investigated the gender effects for prevalence and perception of SH have been relatively few thus far 19, 23). Therefore, the aim of this crosssectional study was to investigate the situation with regards to SH among staff in two general hospitals in Taiwan. We determined the incidence rates of SH among these staff in the past three months, the frequency of each type and the perception of SH perpetrated by coworkers, patients and patients families, respectively. Furthermore, we were interested in whether or not there are gender differences for these issues. Method The hospital and study population The study population was enrolled from the largest general teaching hospital in the Northeastern area of Taiwan and its branch. The main hospital provides medical services with a broad range of specialists expected to be on staff in a general hospital (e.g., internal medicine, surgery, emergency medicine). The main services of the branch are focused on psychiatry, rehabilitation and oncology. They have a total of 1,162 beds, and there are 1,583 staff working in these two hospitals. The staff in the hospitals include 595 nurses, 263 doctors, 295 medical technicians, 289 administration staff and 141 staff in other specialties (e.g., research assistants, fellowship workers). The overall staff were considered as our target study population. Measurement: The Hospital Sexual Harassment Questionnaire (HSHQ) The questionnaire used in this study was developed after consulting the Index of Sexual Harassment (ISH) 28), and we modified the questionnaire to better suit hospital circumstances. For example, the perpetrators of SH, which were peers or supervisors in the ISH, were replaced with coworkers, patients and patients families, respectively. In addition, a five-point Likert scale was used instead of the seven-point one in the ISH. There were four experts who joined together to develop the questionnaire, including a psychologist with a PhD degree whose major specialty was psychological measurement, a hospital psychiatrist and two professors with specialties in gender relationships and English, respectively. The meeting for questionnaire development was to establish the content validity, logic of the questions and suitability for Taiwanese culture. The questionnaire was self-reported and anonymous. The first page of the questionnaire was administered to capture demographic data, including gender, age, marital status and work specialties. In the next three pages, the questionnaire asked about the individuals

3 58 J Occup Health, Vol. 54, 2012 experience with regards to sexual harassment (SH) at the hospital in the past three months perpetrated by coworkers, patients and patients families. A subscale of Coworker SH (C-SH) was included on page 2, and the first question was a dichotomous one (yes or no): Have you experienced sexual harassment from your coworkers in the past three months? Subsequently, there are 20 items measured on a five-point Likert scale (1, representing never, to 5, representing always) on this page. These items were arranged in order of the empirical severity of sexual harassment circumstances from My coworker(s) tells sex jokes (item 1) to My coworker(s) forced me into having a sexual relationship (item 20). The subscales of Patient SH (P-SH) and Patient s Family SH (F-SH) were included on page 3 and 4, respectively. They also included 20 items and one dichotomous question with the same description as for page 2, but the objects were switched from coworkers to patients (page 3) and patients families (page 4). Due to the anonymous character of the questionnaire, test-retest reliability was not available in our study. The Cronbach s alpha coefficient for the C-SH subscale was 0.68, while those for the P-SH and F-SH subscales were 0.93 and 0.77, respectively. Study procedure The study protocol was approved by the institutional review board (IRB) of the local institute. Due to the sensitive nature of the study purpose, the privacy of participants and each individual s willingness to participate were strictly protected. To recruit study participants, we looked to capture survey data from all of the 1,583 staff members of these two hospitals. From January to April in 2010, the questionnaire (HSHQ) was sent to each department, accompanied by a cover letter explaining the study purpose. In addition, there was a clear statement on the cover letter, which elaborated that the individuals who received this questionnaire could freely decide whether or not they wish to participate in the study. This type of questionnaire should be anonymous, and there was no gift or reward for participating in the study. The head nurse, secretaries or assistants of each department distributed the blank questionnaires to every available member. Meanwhile, we provided a large envelope in each department, so that the volunteering participants could put their answered questionnaires into it by themselves. These questionnaires were subsequently collected by staff members of each department within one week. We issued a total of 1,235 questionnaires, and 630 answered questionnaires were returned (51.0%). After invalid questionnaires were excluded, 536 respondents were available for analysis (the overall response rate was 43.4%). Statistical analysis Basic descriptive statistics were calculated to determine the distributions of the demographic characteristics and frequency of SH among the study population. Chi-square tests were used to compare gender differences in demographics and experiences of SH by coworkers, patients and patients families. Due to the small numbers of subjects with SH experiences, the variables of age and work specialty were redivided into dichotomous categories (<30 yr old vs. 31 yr old; nurse vs. non-nurse). Logistic regression analysis was then applied to investigate the independent effect of each dichotomous demographic variable on C-SH, P-SH and F-SH, respectively. Questionnaires with missing data for gender and questionnaires in which the C-SH, P-SH or F-SH subscale was left completely blank were excluded from the analyses (N=94). Because we analyzed the SH circumstances item by item, the few missing responses among the items in the HSHQ were not imputed by other values. The response of each item in the HSHQ was recategorized into a dichotomous variable. The answer of never in the Likert scale was recoded as non-observed, and the responses from rarely to always were recoded as observed. Because female nurses were the most dominant population in this study, the analyses for them were performed separately. Chi-square tests were applied to compare gender differences in distributions of sexual harassment by coworkers, patients and patients families. To determine the differences in perceptions of SH among genders, some items in the HSHQ, which showed a higher incidence or difference between genders, were taken for further analysis. The contingency coefficient was applied to examine the possible correlation of the sexual harassment experience with responses for these items by gender. All statistical analyses were performed with SPSS software version 16.0, and a p-value <0.05 was considered to be statistically significant. Results Among the 536 respondents, 448 (83.6%) were female and 88 (16.4%) were male. There were significant differences in distributions of age, marital status and work positions between genders. In comparing the female and male respondents, the female respondents were found to be younger, not married, and to have the work specialty of being a nurse. The demographic characteristics and experiences of SH by gender are summarized in Table 1. Of the total respondents, there were 13 (2.4%),

4 Liang-Jen WANG, et al.: Sexual Harassment in General Hospitals 59 Table 1. Characteristics of the study participants and their experiences with regards to sexual harassment in the past three months and comparison between genders Total (n=536) (n=448) (n=88) Chi-Square p value Age (Y/O) 31.57*** < (44.3) 213 (89.9) 24 (10.1) < (36.1) 160 (82.9) 33 (17.1) (15.0) 63 (78.8) 17 (21.2) >51 25 (4.6) 12 (48.0) 13 (52.0) Marital status 8.84** Married 234 (43.7) 183 (78.2) 51 (21.8) Single 296 (55.2) 260 (87.8) 36 (12.2) Missing data 6 (1.1) 5 (84.3) 1 (16.7) Specialty 168.3*** Nurse 250 (46.6) 248 (92.2) 2 (0.8) <0.001 Doctor 45 (8.4) 8 (17.8) 37 (82.2) Administration staff 118 (22.0) 93 (78.8) 25 (21.2) Technician 101 (18.8) 81 (80.2) 20 (19.8) Others 18 (3.4) 14 (77.8) 4 (22.2) Missing data 4 (0.7) 4 (100.0) 0 (0.0) SH by coworker 0.39 Yes 13 (2.4) 10 (76.9) 3 (23.1) No 513 (95.7) 428 (83.4) 85 (16.6) Missing data 10 (1.9) 10 (100.0) 0 (0.0) SH by patient 1.00 Yes 23 (4.3) 21 (91.3) 2 (8.7) No 507 (94.6) 423 (83.4) 84 (16.6) Missing data 6 (1.1) 4 (66.7) 2 (33.3) SH by patient s family 0.19 Yes 9 (1.7) 8 (88.9) 1 (11.1) No 515 (96.1) 430 (83.5) 85 (16.5) Missing data 12 (2.2) 10 (83.3) 2 (16.7) *, ** and ***: p<0.05, p<0.01 and p<0.001 for comparison between genders, respectively. SH: sexual harassment. 23 (4.3%), and 9 (1.7%) subjects reporting that they experienced sexual harassment by their coworkers (C-SH), patients (P-SH), and patients families (F-SH) in the past three months, respectively. There were no significant differences with these experiences between female and male groups. For female nurses alone (n=248), the subjects who experienced C-SH, P-SH and F-SH were 7 (2.8%), 14 (5.6%) and 7 (2.8%), accordingly. Table 2 presents the results of a logistic regression analysis of the independent effects of demographic variables on C-SH, P-SH and F-SH, respectively. The results showed that subjects who had experienced C-SH were associated with an age older than 31 yr (adjusted OR=5.46, p=0.022). No significant factors associated with P-SH and F-SH were found in our study population. The frequency distributions of each item in the C-SH, P-SH and F-SH are displayed in Table 3, by gender. For both genders, telling sex jokes was the most common situation in C-SH, P-SH and F-SH. Among the circumstances of C-SH, the second most commonly occurring form was describing me with sexually explicit terminology for the female group. For males, the items displaying pornographic photographs or pictures and describes me with sexually explicit terminology were tied, with the second highest incidence rates. Among the circumstance of P-SH and F-SH, the sequences of positive response frequency in items were similar. The second most common circumstance was gesture or stares at me in a sexual manner for the female group, and the second most common circumstance in the male group was

5 60 J Occup Health, Vol. 54, 2012 Table 2. Multivariate-adjusted odd ratios (ORs) of each variable for sexual harassment by a coworker, patient or patient s family estimated by logistic regression analyses Coworker Patient Patient s family Adj. odds ratio (95% CI) p value Adj. odds ratio (95% CI) p value Adj. odds ratio (95% CI) p value Gender (Reference) (Reference) (Reference) 0.48 (0.10, 2.27) (0.32, 7.48) (0.15, 21.15) Age <30 y/o (Reference) (Reference) (Reference) 31y/o 5.46 (1.29, 23.24) 0.022* 0.66 (0.22, 2.00) (0.64, 12.80) Marital status Married (Reference) (Reference) (Reference) Single 2.60 (0.74, 9.12) (0.40, 3.98) (0.59, 20.18) Specialty Nurse (Reference) (Reference) (Reference) Others 0.45 (0.12, 1.71) (0.29, 1.97) (0.03, 1.36) *p value<0.05. Adj. odds ratio: adjusted odds ratio. Table 3. Frequency distributions of experience for each item describing the circumstances of sexual harassment by a coworker, patient or patient s family in the past three months Perpetrator Coworker Patient Patient s family Item of hospital sexual harassment questionnaire (n=448) (n=88) (n=448) (n=88) (n=448) (n=88) 1. Tells sex jokes 292 (65.2) 60 (68.2) 138 (30.8)*** 10 (11.4) 71 (15.8)* 5 (5.7) 2. Displays pornographic photographs or pictures 17 (3.8) 8 (9.1) 15 (3.3) 3 (3.4) 7 (1.6) 2 (2.3) 3. Produces sexually explicit graffiti for display 5 (1.0) 1 (1.1) 9 (2.0) 1 (1.1) 2 (0.4) 0 (0.0) 4. Distributes porn videos 5 (1.0) 7 (8.0)*** 5 (1.1) 1 (1.1) 3 (0.7) 0 (0.0) 5. Sends sexually explicit s to me 17 (3.8) 8 (9.1) 3 (0.7) 1 (1.1) 2 (0.4) 0 (0.0) 6. Sends sexually explicit letters, cards, or articles to me 13 (2.9) 3 (3.4) 7 (1.6) 1 (1.1) 3 (0.7) 0 (0.0) 7. Describes me with sexually explicit terminology 36 (8.0) 8 (9.1) 27 (6.0) 2 (2.3) 4 (0.9) 0 (0.0) 8. Creates sexually offensive rumors with my appearance or body 22 (4.9) 4 (4.5) 13 (2.9) 1 (1.1) 4 (0.9) 0 (0.0) 9. Uses implications to indicate my sexual behavior or activity 18 (4.0) 6 (6.8) 16 (3.6) 1 (1.1) 8 (1.8) 0 (0.0) 10. Performs a gesture or stares at me in a sexual manner 14 (3.1) 5 (5.7) 33 (7.4)* 1 (1.1) 19 (4.2)* 0 (0.0) 11. Places obscene phone calls to me 8 (1.8) 1 (1.1) 18 (4.0) 1 (1.1) 7 (1.6) 0 (0.0) 12. Unwanted, persistent invitations 8 (1.8) 0 (0.0) 13 (2.9) 2 (2.3) 5 (1.1) 0 (0.0) 13. Stalks me and asks for a relationship 4 (0.9) 1 (1.1) 9 (2.0) 1 (1.1) 3 (0.7) 0 (0.0) 14. Unpleasant, unwanted physical touching 16 (3.6) 2 (2.3) 31 (6.9)* 1 (1.1) 9 (2.0) 0 (0.0) 15. Attempts to block my pathway and forces physical contact 6 (1.3) 3 (3.4) 16 (3.6) 1 (1.1) 7 (1.6) 0 (0.0) 16. Touches him/herself in a sexually explicit way in front of me 1 (0.2) 1 (1.1) 13 (2.9) 1 (1.1) 2 (0.4) 0 (0.0) 17. Touches me in a sexually offensive manner 3 (0.7) 2 (2.3) 16 (3.6) 1 (1.1) 3 (0.7) 0 (0.0) 18. Offers of advantages in exchange of sexual favors 1 (0.2) 0 (0.0) 3 (0.7) 1 (1.1) 1 (0.2) 0 (0.0) 19. Threatens me for sexual favors 1 (0.2) 0 (0.0) 2 (0.4) 1 (1.1) 1 (0.2) 0 (0.0) 20. Forces me into having a sexual relationship 2 (0.4) 0 (0.0) 2 (0.4) 0 (0.0) 1 (0.2) 0 (0.0) *, ** and ***: p<0.05, p<0.01 and p<0.001 for comparison between genders (significantly higher), respectively. The figures in bold-face type represent the items with the highest and second highest incidence rates in each perpetrator category. displaying pornographic photographs or pictures followed by telling sex jokes. There were significant differences in several circumstances between genders, and these results are also shown in Table 3. Compared with the female group, the males had a higher opportunity to be exposed to porn videos by their coworkers. Meanwhile, the females were more frequently exposed to sex jokes and sexual remarks by patients and their families, and unwanted physical touching by patients in the workplace, as compared with male workers. In terms of the perception of SH, there were some significant differences between genders as outlined in Table 4. In summary, if the sex jokes were told by

6 Liang-Jen WANG, et al.: Sexual Harassment in General Hospitals 61 Table 4. Correlation coefficients for respondents experience with items and their recognition of sexual harassment by a coworker, patient or patient s family in the past three months (n=438) Coworker Patient Patient s family (n=88) (n=444) (n=86) (n=438) (n=86) Tells sex jokes *** *** 0.03 Displays pornographic photographs and pictures 0.13** *** *** 0.02 Distributes porn videos 0.13** *** - a a Sends sexually explicit s to me *** - a a Describes me with sexually explicit terminology *** 0.49*** ** - a Performs a gesture or stares at me in a sexual manner 0.23*** 0.22* 0.39*** - a 0.22*** - a Unpleasant, unwanted physical touching 0.38*** 0.81*** 0.52*** - a 0.34*** - a : Contingency coefficient. *, ** and ***: p<0.05, p<0.01 and p<0.001 for the contingency coefficient. a : No statistics were computed because all case rated the same value. coworkers, it was not viewed as SH in both genders. However, if sex jokes were told by patients or their families, they were regarded as SH among females, but not among males. For the circumstance of displaying pornographic photographs or pictures, females regarded it as SH for all the three perpetrators, but males did not. For the item of describing me with sexually explicit terminology, females regarded it as SH among patients and their families, but not among coworkers. In contrast, males did view it as SH among coworkers, but not among patients. Discussion In this study, we reported the three-month incidence of sexual harassment (SH) among workers in two general hospitals in Taiwan. The respondents who were over 31 yr old had a higher rate of SH by their coworkers. Telling sex jokes was the most common type of SH reported by respondents. The males had higher opportunities to be exposed to porn videos by their coworkers. The females more frequently suffered from sex jokes and sexual remarks by patients and patients families, and unwanted physical touching by patients in the workplace. In addition, there were significant differences with regards to perception of SH between genders in several circumstances of SH. The three-month incidence rates of SH by coworkers, patients and patient s families in our study population were 2.4, 4.3, and 1.7%, respectively. These values were lower than most reports in previous international studies, whereas the previous studies investigated the annual or lifetime prevalence. For the studies focused on nurses, some of them reported that physicians or other coworkers were identified as the primary perpetrators of SH 7, 18, 29). Other literature demonstrated that the most common perpetrators in all types of SH were patients, followed by doctors or other staff 12, 25). The prevalence of SH by patients relatives was reported to be lower than the prevalences of SH by colleagues and patients 7, 18). The differences in the sources of SH might reflect the power imbalances and cultural variations between hospital circumstances in countries, and the results in our study revealed that patients were the most common perpetrators of SH. Surprisingly, respondents who were older than 31 yr old had a higher rate of SH by colleagues, as compared with younger staff. This is opposite to the empirical assumption that younger workers might be more sexually attractive and more prone to being harassed 30). Among hospital workers, Kinard (1995) reported that the risk of being sexually harassed was higher for female workers in jobs where the vast majority of their coworkers are of the opposite sex. In studies of nurses, higher risks of SH were found among younger nurses compared with older nurses 31). However, Celik (2007) demonstrated that being single and increased work experience were important determinants of being sexually harassed at the workplace. Nevertheless, these studies investigated the prevalence rate rather than incidence rate, and it is reasonable to assume that a longer time being at work simply meant a longer amount of time for exposure to SH. Because the age effect was only found in C-SH in our study, it warrants further investigation as to whether the older workers are more familiar with their coworkers and can be more easily sexually harassed by them. The types of SH were variable for different perpetrators in international studies. Researchers found the most prevalent SH behaviors towards nurses include unwanted sex jokes, stories, questions or words (83.5%) 18), sexual remarks (33%), touching (26%) and pressure for dates (15%) 29). For respondents who were doctors or medical students, sex jokes, sugges-

7 62 J Occup Health, Vol. 54, 2012 tive looks and sexual remarks were the most prevalent types of SH 13, 15). Nurses in Turkey report the most common SH event was exposure of parts of the body in a sexually suggestive way by patients; suggestive looks, sexual teasing, jokes, remarks or questions from doctors; and pressure to date by patients relatives 7). When the occupation or workplace is male dominated, in terms of power, there is a greater chance of sexual harassment for a woman working in that environment 23). The results in our study demonstrated that the severe types of SH seldom occurred, especially for male respondents perpetrated by patients or patients families. However, it would be interesting to clarify whether the experiences of SH were associated with properties of their work in a future investigation. The results in our study did not reveal gender differences for the overall incidence rate of SH; however, there were several items that showed differences between genders. The males were more often noted to be exposed by their coworkers to porn videos, although they seemed not to regard it as SH. The female subjects in our study reported that they more frequently suffered from sex jokes and sexual remarks by patients and patients families, and they did perceive these situations as SH. In terms of attitude towards SH, the behavior of uninvited sexual teasing, jokes and remarks by either staff or patients was classified as SH among females more so than among males 25). Bronner (2003) reported that men felt significantly fewer negative feelings following teasing remarks or sex jokes than women 19). SH behavior may be seen as appropriate and normal behavior to some men, and female nurses usually perceived all types of SH as being more severe than male nurses 22, 23). Gender typing at an early age influences attitudes toward SH. The results of our study showed the gender differences for perceptions toward several circumstances of SH from various perpetrators. Nevertheless, the number of subjects who experienced SH was very small, especially for male respondents. Thus, the gender differences should be further examined in future studies with larger sample sizes. This study has some limitations. First, the subjects were not randomly selected, and the reasons for not participating or response to questions are not known. As a result, the findings do not represent the total population of hospital workers in Taiwan. Furthermore, the characters of these two hospitals are quite different, and the risks of SH might be variable between departments providing medical services. However, information about the departments of the respondents was not captured in this study; thus, it was unavailable for analysis of the differences. Second, our data are solely self-reported, and the reliability and validity should be further examined in future studies. Although the questionnaires were anonymous, we cannot rule out all possibility of biased estimates. Besides, the questionnaire aimed to capture the incidence of SH in three months. Our original concern was to eliminate the recall bias from memory confusion; however, it also reduced the number of subjects reporting experience of SH during this short period. The small number further limited the statistical power to determine the possible risk factors of SH, as well as gender differences. Third, there were a certain number of questionnaires with missing data among subscales, and an information bias may result from data being missing from the questionnaires. Fourth, we determined the perception of SH simply by statistical correlation of the responses for items, rather than asking the respondents how they felt directly. This might result in somewhat of a misinterpretation of the results. Finally, the number of male respondents was not large, and the serious types of SH rarely occurred among them. Thus, there were no valid cases for detecting the gender differences for perception of severe types of SH. Gruber (1990) reported that the incidence of SH might be influenced by several methodological problems, such as response rate, sample size, sample diversity, and instrument construction 32). However, due to the sensitive and embarrassing properties of the study topic, this kind of study is difficult to perform without the study design of an anonymous and self-reporting survey in Taiwan. Conclusions Despite these limitations, the strength of this study is that it contributes incidence rates of SH in two general hospitals in Taiwan, involving the staff of various specialties. We also compared the differences for frequencies of SH by various perpetrators between genders, as well as the differences in their perceptions. The disclosure of SH is still somewhat taboo in many countries. The Sexual Harassment Prevention Law was established in 2006 in Taiwan in order to protect the rights of SH victims. This study provides a preliminary reference for fellow medical administrators. It might contribute to recognition of the types of SH commonly occurring in hospitals, and further assist in establishing education plans and policies to prevent SH and possible misunderstandings between genders. Such a proactive approach to address SH will also ensure that patients receive the best care possible from those individuals who work in a hospital environment.

8 Liang-Jen WANG, et al.: Sexual Harassment in General Hospitals 63 Acknowledgments: This work was supported by the Chang-Gung Memorial Hospital Research Project (CMRP C). The authors also wish to express their gratitude to Dr. Liu KH (the Special Adjunct Professor, Department of Counseling Psychology, Chinese Culture University) and Professor Lwo LS (Institute of Education & Dean of the College of Humanities and Social Sciences, National Taiwan Ocean University) for establishing the Hospital Sexual Harassment Questionnaire. References 1) Crull P, Cohen M. Expanding the definition of sexual harassment. Occup Health Nurs 1984; 32: ) Fitzgerald LF. Sexual harassment. Violence against women in the workplace. Am Psychol 1993; 48: ) Takaki J, Tsutsumi A, Fujii Y, et al. Assessment of workplace bullying and harassment: reliability and validity of a Japanese version of the negative acts questionnaire. J Occup Health 2010; 52: ) Chaudhuri P. Experiences of sexual harassment of women health workers in four hospitals in Kolkata, India. Reprod Health Matters 2007; 15: ) Luegenbiehl DL, Goldbort JI. Sexual harassment in the hospital. Revolution 1993; 3: 64 5, 95. 6) Hibino Y, Ogino K, Inagaki M. Sexual harassment of female nurses by patients in Japan. J Nurs Scholarsh 2006; 38: ) Kisa A, Dziegielewski SF, Ates M. Sexual harassment and its consequences: a study within Turkish hospitals. J Health Soc Policy 2002; 15: ) Di Martino V. Workplace violence in the health sector-country case studies Brazil, Bulgaria, Lebanon, Portugal, South Africa, Thailand, plus an additional Australian study: synthesis report. Geneva: ILO/ICN/WHO/PSI Joint Programme on Workplace Violence in Health Sector, ) Chen WC, Hwu HG, Kung SM, Chiu HJ, Wang JD. Prevalence and determinants of workplace violence of health care workers in a psychiatric hospital in Taiwan. J Occup Health 2008; 50: ) Tammelleo AD. Nurse sues hospital for sexual harassment. Regan Rep Nurs Law 1993; 34: 1. 11) Hamlin L, Hoffman A. Perioperative nurses and sexual harassment. AORN J 2002; 76: ) Cogin J, Fish A. Sexual harassment a touchy subject for nurses. J Health Organ Manag 2009; 23: ) Phillips SP, Schneider MS. Sexual harassment of female doctors by patients. N Engl J Med 1993; 329: ) Nagata-Kobayashi S, Maeno T, Yoshizu M, Shimbo T. Universal problems during residency: abuse and harassment. Med Educ 2009; 43: ) Larsson C, Hensing G, Allebeck P. Sexual and gender-related harassment in medical education and research training: results from a Swedish survey. Med Educ 2003; 37: ) Kinard JL, McLaurin JR, Little B. Sexual harassment in the hospital industry: an empirical inquiry. Health Care Manage Rev 1995; 20: ) Chen WC, Sun YH, Lan TH, Chiu HJ. Incidence and risk factors of workplace violence on nursing staffs caring for chronic psychiatric patients in taiwan. Int J Environ Res Public Health 2009; 6: ) Celik Y, Celik SS. Sexual harassment against nurses in Turkey. J Nurs Scholarsh 2007; 39: ) Bronner G, Peretz C, Ehrenfeld M. Sexual harassment of nurses and nursing students. J Adv Nurs 2003; 42: ) Fitzgerald LF, Gelfand, M.J., Drasgow, F. Measuring sexual harassment: theoretical and psychometric advances. Basic Appl Social Psychol 1995; 17: ) Terpstra DE, Baker DD. A hierarchy of sexual harassment. J Psychol 1987; 121: ) Madison J, Minichiello V. Recognizing and labeling sex-based and sexual harassment in the health care workplace. J Nurs Scholarsh 2000; 32: ) Fiedler A, Hamby E. Sexual harassment in the workplace: nurses perceptions. J Nurs Adm 2000; 30: ) Hemmasi M, Graf LA, Russ GS. Gender-related jokes in the workplace: sexual humor or sexual harassment? J Appl Social Psychol 1994; 24: ) Morgan JF, Porter S. Sexual harassment of psychiatric trainees: experiences and attitudes. Postgrad Med J 1999; 75: ) Chuang SC, Lin HM. Nurses confronting sexual harassment in the medical environment. Stud Health Technol Inform 2006; 122: ) Chen WC, Hwu HG, Wang JD. Hospital staff responses to workplace violence in a psychiatric hospital in Taiwan. Int J Occup Environ Health 2009; 15: ) Decker AL, Hudson WW. Index of Sexual Harassment. Handbook of Sexuality-Related Measures. Thousand Oaks (CA): Sage; ) Valente SM, Bullough V. Sexual harassment of nurses in the workplace. J Nurs Care Qual 2004; 19: ) Chou KR, Lu RB, Mao WC. Factors relevant to patient assaultive behavior and assault in acute inpatient psychiatric units in Taiwan. Arch Psychiatr Nurs 2002; 16: ) Hibino Y, Hitomi Y, Kambayashi Y, Nakamura H. Exploring factors associated with the incidence of sexual harassment of hospital nurses by patients. J Nurs Scholarsh 2009; 41: ) Gruber JE. Methodological problems and policy implications in sexual harassment research. Popul Res Policy Rev 1990; 9: